出版社：The Institute of Image Information and Television Engineers
摘要：We propose a method for fusing multiple images into a single one. Our method is an extension of a technique for enhancing the contrast of a single image to multiple ones. First, we calculate the logarithm of pixel values in component images and smooth them using a bilateral filter. The smoothed image is the basic component and the difference between the original image and its basic component is the detailed component. We then compress the basic component and add it to the detailed component. Finally we recover the pixel values with an exponential function to obtain contrast-enhanced images. When adding detailed components together, we weigh both components according to their magnitude to preserve salient information in both component images. We tested this method with flash shot, multiple exposure, multiple focus, multiple light and multimodal medical images.
